Adaptive Neuro-Fuzzy Inference System (ANFIS) Real Time Based Power Factor Control by Synchronous Machine Dr. Maamoon F. Al-Kababji Mr. Ahmed Nasser B. Al-Sammak BSc (Eng), PhD Electrical Engineering Department BSc (Eng), Msc Electrical Engineering Department University of Mousl University of Mousl Mousl-Iraq E-mail: al_kababjie@yahoo.com Mousl-Iraq E-mail: Ahmed_Alsammak2003@yahoo.com Abstract : In this paper, the application of an ANFIS technique is used to compensate reactive power of load by controlling the excitation system of synchronous machine. This fuzzy logic based controlled excitation system can thus give a very fast response of the system to meet the required load reactive power and hence keeping the load bus at constant predetermined power factor (PF) value. The over or under compensation, as well as time delay could therefore be eliminated with such a control configuration. It is concluded that the reactive power compensation system with a fuzzy logic controlled synchronous machine is reliable, sensitive, economical, faster and more efficient than an other one with capacitor groups. Matlab7.4 (R2007a) was adopted for the architecture and learning procedure underlying ANFIS, which is a fuzzy inference system implemented in the framework of adaptive networks. By using a hybrid learning procedure, the proposed ANFIS can construct an input-output mapping based on both human knowledge (in the form of fuzzy if-then rules) and stipulated input-output data pairs. In the simulation, the ANFIS architecture is employed to model PF control. A suitable model of the synchronous machine was also presented in this paper. This model was tested with some cases such as sudden change in load, field voltage as well as loading or no loading condition. The basic equations of the motor are also developed in a manner, suitable for the Matlab-Simulink application. The computer and practical results were represented and compared. Good agreement between these results has been observed, which proves the validity of the simulation method used in this work.The variable DC voltage based excitation field current controller was built based on fuzzy logic controller (FLC) to generate the firing angle (alpha) of six-pulse rectifier circuit, to give this voltage changes as the output of the FLC changes depending on the set point (load bus reference PF). Keywords: ANFIS, Fuzzy Logic, Synchronous Machine, Voltage Stability, Excitation Control, Power Factor, Reactive Power Compensation and Matlab-Simulink. The First Electrical Engineering Conference , EEC 07, Faculty of Electrical & Electronic Engineering June 26-28, 2007, Aleppo, Syria 1